INVITED PAPER Cognitive Radio Architecture Evolution Systems for automatic selection of radio bands and operating modes are evolving to meet user needs at specific times and places. By Joseph Mitola, III, Senior Member IEEE ABSTRACT | The radio research community has aggressively embraced cognitive radio for dynamic radio spectrum man- agement to enhance spectrum usage, e.g., in ISM bands and as secondary users in unused TV bands, but the needs of the mobile wireless user have not been addressed as thoroughly on the question of high quality of information (QoI) as a func- tion of place, time, and social setting (e.g. commuting, shop- ping, or in need of medical assistance). This paper considers the evolution of cognitive radio architecture (CRA) in the context of motivating use cases such as public safety and sentient spaces to characterize CRA with an interdisciplinary perspective where machine perception in visual, acoustic, speech, and natural language text domains provide cues to the automatic detection of stereotypical situations, enabling radio nodes to select from among radio bands and modes more intelligently and enabling cognitive wireless networks to deliver higher QoI within social and technical constraints, made more cost effec- tive via embedded and distributed computational intelligence. KEYWORDS | Architecture; cognitive radio; quality of informa- tion (QoI); software defined radio (SDR) I. INTRODUCTION When introduced in 1998–1999 [1], [2], cognitive radio emphasized enhanced quality of information (QoI) for the user, with spectrum agility framed as a means to an end and not as an end in itself. The first research prototype cognitive radio (CR1), for example, learned to turn on Bluetooth to exchange business cards wirelessly when the user’s speech dialogVsensed via (simulated) speech recognitionVexhibited the characteristics of a prototypi- cal setting of Bintroductions,[ meeting new people. This intelligent agent embedded in the personal digital assistant (PDA) was unique in that it had not been programmed to do this, but rather learned this behavior via case-based reasoning (CBR) from its user’s prior manual exchange of electronic business cards. CR1 associated the user’s prior manual use of the PDA’s Bluetooth radio to exchange business cards with cues in the speech domain such as phrases like, BMay I introduce,[ and BVery pleased to meet you.[ CR1 thus synthesized a CBR template to [ GPower-up Bluetooth/>, GExchange Business-cards/>, GPower-down Bluetooth/>] autonomously and could learn by being told (not via rules previously programmed into it) etiquettes for sharing radio spectrum with legacy radios. Intelligent agents like CR1 observe the environment in which they are embedded in order to learn to formulate plans and execute actions that respond intelligently to the user in the environment. These early contributions stimulated many ideas for cooperative spectrum sharing [3], capturing the imagination. During the past five years, the world’s radio research and engineering communities have been developing soft- ware defined radio (SDR) and cognitive radio (CR) for dy- namic radio spectrum sensing, access, and sharing [4]–[6], revealing many regulatory, business, market, and open architecture needs implicit in the broad potential that cognitive radio architecture (CRA) introduces 1 [7]. Radio architecturesVfrom wearable nodes and radio access points to the larger converged networksVhave evolved from the niche market of single-band single-mode car phones of the 1970s to today’s ubiquitous multiband- multimode fashion statements. This paper characterizes architecture evolution, including the near-term multime- dia heterogeneous networks that converge traditional cel- lular architectures with Internet hot spots. This paper also Manuscript received November 11, 2008. Current version published April 15, 2009. The author is with the Charles V. Schafer Schools of Engineering and Science and the School of Systems and Enterprises, Stevens Institute of Technology, Hoboken, NJ 07030 USA (e-mail: [email protected]). Digital Object Identifier: 10.1109/JPROC.2009.2013012 1 See www.sdrforum.org/CRWG. 626 Proceedings of the IEEE | Vol. 97, No. 4, April 2009 0018-9219/$25.00 Ó2009 IEEE
16
Embed
INVITED PAPER CognitiveRadio ArchitectureEvolutionstaff.utar.edu.my/gohhg/research/files2/cognitive... · CognitiveRadio ArchitectureEvolution Systems for automatic selection of radio
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
INV ITEDP A P E R
Cognitive RadioArchitecture EvolutionSystems for automatic selection of radio bands and operating modes are
evolving to meet user needs at specific times and places.
By Joseph Mitola, III, Senior Member IEEE
ABSTRACT | The radio research community has aggressively
embraced cognitive radio for dynamic radio spectrum man-
agement to enhance spectrum usage, e.g., in ISM bands and as
secondary users in unused TV bands, but the needs of the
mobile wireless user have not been addressed as thoroughly
on the question of high quality of information (QoI) as a func-
tion of place, time, and social setting (e.g. commuting, shop-
ping, or in need of medical assistance). This paper considers the
evolution of cognitive radio architecture (CRA) in the context of
motivating use cases such as public safety and sentient spaces
to characterize CRA with an interdisciplinary perspective
where machine perception in visual, acoustic, speech, and
natural language text domains provide cues to the automatic
detection of stereotypical situations, enabling radio nodes to
select from among radio bands and modes more intelligently
and enabling cognitive wireless networks to deliver higher QoI
within social and technical constraints, made more cost effec-
tive via embedded and distributed computational intelligence.
KEYWORDS | Architecture; cognitive radio; quality of informa-
tion (QoI); software defined radio (SDR)
I . INTRODUCTION
When introduced in 1998–1999 [1], [2], cognitive radioemphasized enhanced quality of information (QoI) for the
user, with spectrum agility framed as a means to an end
and not as an end in itself. The first research prototype
cognitive radio (CR1), for example, learned to turn on
Bluetooth to exchange business cards wirelessly when the
user’s speech dialogVsensed via (simulated) speech
recognitionVexhibited the characteristics of a prototypi-cal setting of Bintroductions,[ meeting new people. This
intelligent agent embedded in the personal digital assistant
(PDA) was unique in that it had not been programmed to
do this, but rather learned this behavior via case-based
reasoning (CBR) from its user’s prior manual exchange of
electronic business cards. CR1 associated the user’s prior
manual use of the PDA’s Bluetooth radio to exchange
business cards with cues in the speech domain such asphrases like, BMay I introduce,[ and BVery pleased to meet
you.[ CR1 thus synthesized a CBR template to [ GPower-up
Digital Object Identifier: 10.1109/JPROC.2009.20130121See www.sdrforum.org/CRWG.
626 Proceedings of the IEEE | Vol. 97, No. 4, April 2009 0018-9219/$25.00 �2009 IEEE
looks ahead towards the establishment of sentient spaces[8]–[10], integrated wireless environments that merge
wireless technologies with increasing interplay of radio
engineering with related information services of computer
vision [11] and human language technologies (HLTs) [12].
II . ORGANIZATION
This paper first reviews the concept of architecture inSection III, including prototypical architectures for dynamic
spectrum and embedded agents. It then describes the
apparent lack of a comprehensive metalevel architecture for
distributed heterogeneous networks and their related meta-
level superstructures, including regulatory rule-making and
spectrum auctions. Section IV characterizes the changes in
use case that drive wireless architecture, showing how the
historically significant striving for ubiquity and high datarate is beginning to give way to evolved value propositions in
which appropriately high quality of service (QoS) is merely
the starting point for QoI. Section V therefore develops the
potential for greater integration of cross-discipline infor-
mation sources like video surveillance and human language
technology in future cognitive radio architectures. To help
guide this evolution, QoI is characterized along its several
dimensions in Section VI, while Section VII offers a reviewof challenges and opportunities before the conclusion of
Section VIII.
III . COGNITIVE RADIO ARCHITECTURES
Radio architecture is a framework by which evolving
families of components may be integrated into an evolving
sequence of designs that synthesize specified functionswithin specified constraints (design rules) [13]. A powerful
architecture facilitates rapid, cost-effective product and
service evolution. An open architecture is available to the
public, while a proprietary architecture is the private in-
tellectual property of an organization, government entity,
or nonpublic consortium. Fig. 1 illustrates functional com-
ponents integrated to create an SDR device, which may be
wearable, mobile, or a radio access point in a largernetwork.
The set of information sources of Fig. 1 includes speech,text, Internet access, and multimedia content. Today’s
commercial radio-frequency (RF) channel sets have typically
four chip sets [GSM 900, GSM1800, code-division multiple
access (CDMA), and Bluetooth, for example], evolving in
the near term to a dozen band-mode combinations with
smart antennas and multiple-input multiple-output (MIMO)
emerging [15]. In addition, a channel-set may include a
cable interface to the public switched telephone network(PSTN) (IP or SDH) as well as a radio access point. Any
functions may be null in any realization, eliminating the
related components and interfaces from a given product for
product tailoring and incremental evolution.
With the continued progress of Moore’s law, increas-
ingly large fractions of such functionality are synthesized in
chipsets with software-definable parameters; in the field-
modifiable firmware of field-programmable gate arrays(FPGAs); in software for niche instruction set architectures
(e.g., digital signal-processing chips); and increasingly on
blade server(s) and single-chip arrays of general-purpose
processors like IMEC Belgium’s SIMD4 [16].
Today’s SDRs often are synthesized from reusable code
bases of millions of lines of code [17], the deployment,
management, and maintenance of which poses configura-
tion challenges. The SDR software typically is organized asradio applications objects layered upon standard infra-
structure software objects for distributed processing such
as the SDR Forum’s software communications architecture
(SCA),2 which originally was based primarily on CORBA3
[18]. The Object Management Group’s evolved SCA has a
platform independent model with platform specific models
for software-based communications.4 Infrastructure layers
of such architectures are illustrated in the larger context ofFig. 2. Prior to circa 2005, such architecture was overkill
for handsets, but radio access networks have grown to
millions of lines of code consisting of the kinds of software
objects with the types of layering illustrated in the figure,
and now applicable to handsets and systems on chip (SoC).
As Mahonen (RWTH, Aachen, Germany) was among the
first to clearly differentiate [19], software radio and cogni-
tive radio are Binterlinked and are family members, butthey also have distinctive roadmaps[ as the evolution of
cognitive radio architecture illustrates. Again from
Mahonen, BThere are still formidable hardware and algo-
rithm development problems (such as AD/DA-converters
. . .) before full (ideal) all-in-one software radio can be
built.[ However, Bthe basic paradigm in the cognitive
radios is to provide technologies, which enable radio to
(c) 1998–2006 Mitola’s STATISfaction, reprinted with permission
Fig. 1. Set theoretic model of SDR architecture [14].
2See www.sdrforum.org and www.jtrsjpeo.gov.3www.omg.org/corba.4The development of the SDR Forum’s SCA was sponsored in 1996
by the Defense Advanced Research Projects Agency (DARPA) and theU.S. Air Force towards an industry standard open architecture forSPEAKeasy II evolving to the Joint Tactical Radio System (JTRS) in 1997;as of May 2008, the JTRS program configuration-managed the U.S.Department of Defense configuration of the SDR Forum’s SCA.
Mitola: Cognitive Radio Architecture Evolution
Vol. 97, No. 4, April 2009 | Proceedings of the IEEE 627
reason about its resources, constraints, and be aware of
users/operators’ requirements and context.[What are the resources and constraints? Arguably since
the early 1900s, conventional radio architecture has been
constrained by government regulatory frameworks accurate-
ly characterized as lanes in the road: bands large and small
allocated to specific uses, in the public interest. That
regulatory regime addressed the public interest within theconstraining economics of radio devices and related in-
frastructure (such as large, expensive television broadcast
towers). This was economically efficient (arguably Pareto
efficient) from the Btransistor radio[ and television era to the
deployment of first- and second-generation cellular radio.
However, today’s low-cost multiband multimode wearable
wireless fashion statements; the proliferation of cellular
infrastructure; the gigabit per second core IP networks; andwireless local-area network (WLAN) consumer products
have proliferated wireless access points of all sorts in the
home, workplace, and, seemingly, just about everyplace else
in developed economies. The new wireless ubiquity and
heterogeneity offers rapidly emerging alternatives to the
lanes in the road that include dynamic spectrum access.
A. Dynamic Spectrum AccessBriefly, dynamic spectrum access is the process of in-
creasing spectrum efficiency via the real-time adjustment
of radio resources, e.g., via a process of local spectrum
sensing, probing, and the autonomous establishment of
local wireless connections among cognitive nodes and
networks. As originally proposed, cognitive radio envi-
sioned real-time spectrum auctions among diverse con-
stituencies, using for one purpose, such as cellular radio,
spectrum allocated and in use for another purpose, such as
public safety, and vice versa, in order to multiply the
number of radio access points for public safety and to moreefficiently use public safety spectrum commercially during
peak periods [1]. Although that initial example has yet to
be fully realized, the U.S. Federal Communications Com-
mission encouraged the application of that technology to
the secondary use of underutilized television spectrum,
such as in an ad hoc, short-range WLAN in spectrum that is
allocated to another primary purpose, such as broadcast
television. In addition, the principles of cognitive radio fordynamic spectrum also apply to enhance the efficiency of
use within and across each Blane in the road,[ such as via
the intelligent selection among multiple alternative phy-
sical (PHY) media access control (MAC) layers (alterna-
tive lanes in the spectrum road) by cognition across
network, transport, and application layers of the protocol
stack [20]. Researchers characterize the advantages of
short-term localized dynamic spectrum auctions [21], [22],including rigorous and comprehensive treatments in the
European Community (EC)’s precompetitive End to End
Reconfigurability program [23]. In spite of commercial
(c) 1998–2006 Mitola’s STATISfaction, reprinted with permission
Fig. 2. Software complexity of wireless devices and infrastructure leads to object and API layering [13].
Mitola: Cognitive Radio Architecture Evolution
628 Proceedings of the IEEE | Vol. 97, No. 4, April 2009
proposals [24], only long-term large-capacity anonymousleasing appears to be established in the marketplace.5
The endorsement of the FCC for cognitive radio in
secondary markets offered opportunities for improved
spectrum utilization [25]. In addition, the National Institute
of Information and Communications Technology (NICT),
Yokosuka, Japan, have characterized SDR and cognitive radio
from technical [26], [27] and regulatory [28] perspectives.
Ofcom, the regulatory body of the United Kingdom, remainsappropriately skeptical of the economics of dynamic spectrum
[29]. On the other hand, the Commission for Communica-
tions Regulation, Ireland, imposes constraints [30] but also
encourages innovation such as by allocating over 100 MHz of
spectrum for experiments and demonstrations during the
IEEE Dynamic Spectrum (DySPAN) Conference in 2007 in
Dublin. Guatemala6 employs Titulos de Usurfrucato de
Frecuencias, specifying spectrum use parameters in greatdetail, which establishes a strong reference point for the
liberalization of spectrum allocation towards dynamics [31].
In Europe, countries including Austria, Sweden, and the
United Kingdom apparently have sanctioned de facto
transfers of spectrum rights among spectrum licensees, while
a recent EU Framework Directive empowers all European
Commission countries to introduce secondary trading of
spectrum usage rights [18].The SDR Forum’s CR working group (CRWG) and the
inclusion of CR in its annual academic challenge promotes
the global interchange among academic research and
industry development of cognitive radio in SDR.7 DARPA’s
XG program [32], [33] put substantial emphasis on the near-
term potential of smart radios to share the radio spectrum
dynamically, leading, among other things, to the success of
the IEEE Dynamic Spectrum (DySPAN) conferences in2005 and 2007, where XG research results were demon-
strated [34]. XG simplified the ideal cognitive radio archi-
tecture (iCRA [35]) to a simple rule-engine that controls the
radio’s air interfaces to conform to spectrum use policies
expressed in a rule-based policy language. This yields a
Karlsruhe, Karlsruhe, Germany; TU Delft and Twente,
The Netherlands; University of California at Berkeley;
Virginia Tech; Virtual Center of Excellence Wireless, U.K.;
and Zhejiang University, Hangzhou, China; to all of
whom the author is particularly indebted regarding thisProceedings paper.
RE FERENCES
[1] J. Mitola, III, BCognitive radio,[ Licentiatethesis, KTH, Royal Inst. of Technol.,Stockholm, Sweden, Sep. 1999.
[2] J. Mitola, III, BCognitive radiofor flexible mobile multimediacommunications,[ in Proc. IEEEMobile Multimedia Commun. Conf.(MoMuC), New York, Nov. 1999.
[3] J. Mitola, III, Cognitive Radio Architecture.New York: Wiley, 2006.
[4] F. Fitzek and M. Katz, Cooperation in WirelessNetworks. Dordrecht, The Netherlands:Springer, 2006.
[5] Proc. IEEE Dyn. Spectrum (DySPAN) Conf.2005 and 2007, New York.
[6] Proc. IEEE Cognitive Radio OrientedWireless Netw. (CROWN) Conf., New York,2007.
[7] IEEE Cognitive Radio-Related StandardsActivities: P1900, SCC41, and 802.22.
[8] D. Murley, private communication, 1998.
[9] J. Riihijarvi and P. Mahonen, BExploitingspatial statistics of primary and secondaryusers towards improved cognitive radionetworks,[ in Proc. CrownCom 08, Aachen,Germany, 2008.
[10] E. Meshkova, J. Riihijarvi,P. Mahonen, and C. Kavadias, BModelingthe home environment using ontologywith applications in software configurationmanagement,[ in Proc. 15th Int. Conf.
Telecommun. (ICT), St. Petersburg, Russia,Jun. 2008.
[11] BCognitive vision,[ AAAI Mag., Oct. 2006.
[12] Proc. Human Language Technol. 2007,Stroudsburg, PA, Apr. 2007.
[13] J. Mitola, Software Radio Architecture.New York: Wiley, 2000.
[14] J. Mitola, III, BSoftware radio architecture:A mathematical perspective,[ IEEE J. Sel.Areas Commun., May 1999.
[15] H. Huang, Z. Zhang, P. Cheng, G. Yu, andO. Qui, BThroughput analysis of cognitiveMIMO system,[ in Proc. Int. WorkshopCross Layer Design (IWCLD’07), Jinan, China,Sep. 2007.
Mitola: Cognitive Radio Architecture Evolution
Vol. 97, No. 4, April 2009 | Proceedings of the IEEE 639
[16] D. Novo, B. Bougard, P. Raghavan,T. Schuster, L. Van der Perre (IMEC,Leuven, Belgium), and H.-S. Kimand H. Yang (Samsung). (2006).BEnergy-performance exploration of aCGA-based SDR processor,[ in Proc. SDRForum Tech. Symp. [Online]. Available: www.sdrforum.org
[17] C. Linn and E. Koski, BThe JTRS program:Software-defined radios as a softwareproduct line,[ in Proc. 10th Int. SoftwareProduct Line Conf., New York, Aug. 2006.
[18] J. Mitola, III, Ed., BSoftware radio,[IEEE Commun. Mag., May 1995.
[19] P. Mahonen, BCognitive trends in making:Future of networks,[ in Proc. PIMRC,New York, 2004.
[20] B. S. Manoj, M. Zorzi, and R. Rao, BA newparadigm for cognitive networking,[ in Proc.IEEE DySPAN, New York, 2007.
[21] L. Kovacs and A. Vidacs, BSpectrumauction and pricing in dynamic spectrumallocation networks,[ in Proc. IEEE DySPAN,New York, 2007.
[22] S. Gandhi, C. Buragohain, L. Cao, H. Zheng,and S. Suri, Towards Real-Time DynamicSpectrum Auctions. Santa Barbara, CA:Univ. of California, May 22, 2007.
[23] D. Grandblaise, C. Kloeck, K. Moessner,V. Rodriguez, E. Mohyeldin,M. K. Pereirasamy, J. Luo, and I. Martoyo,BTechnoVEconomic of collaborative basedsecondary spectrum usageVE2R research:Project Outcomes Overview,[ in Proc.IEEE DySPAN, New York, 2005.
[24] Google, Inc., BLetter to the US FCC, ex partevia electronic filing, RE: ex parte filing;service rules for the 690–746, 747–762, and777–792 MHz Bands,[ WC Docket 06-150;WC Docket 06-129; PS Docket 06-229;WT Docket 96-86, May 21, 2007. [Online].Available: www.fcc.org
[25] FCC, BReport and order facilitatingopportunities for flexible, efficient,and reliable spectrum use employingcognitive radio technologies,[ Washington,D.C., Mar. 11, 2005.
[26] H. Harada, BSoftware defined radioprototype toward cognitive radiocommunication systems,[ in Proc.IEEE DySPAN, New York, 2005.
[27] H. Harada, BRegulatory perspective ofJapan,[ in Proc. VCE Regulatory Workshop,London, U.K.
[28] H. Harada, BAdvances in flexible radiotechnology to support cognitive radio,[ inProc. VCE Int. Res. Workshop Intell. SpectrumUsage Pers. Commun., London, U.K.,Apr. 2007.
[29] W. Webb, BKeynote address,[ IEEE DySPAN2007.
[30] K. E. Nolan, E. Ambrose, and D. O’Mahony,BCognitive radio: Value creation andvalue-migration,[ in Proc. SDR ForumTech. Conf. 2006. [Online]. Available:www.sdrforum.org
[31] R. Ercole, BInnovation, spectrum regulation,and DySPAN technologies access tomarkets,[ in Proc IEEE DySPAN, New York,2005.
[32] P. Mannion, BSmart radios stretchspectrum,[ EE Times, Dec. 5, 2005.
[33] F. Seelig, BA description of the August 2006XG demonstrations at Fort A. P. Hill,[ inProc. DySPAN 2007, New York, Apr. 2007.
[34] F. Perich, BPolicy-based networkmanagement for next generationspectrum access control,[ inProc. DySPAN 2007, New York,Apr. 2007.
[35] J. Mitola, III, BCognitive radio: An embeddedagent architecture for software definedradio,[ doctoral thesis, KTH, Royal Inst.of Technol., Stockholm, Sweden,Jun. 2000.
[36] S. Haykin, BCognitive radio:Brain-empowered wirelesscommunications,[ IEEE J. Sel.Areas Commun., Feb. 2005.
[37] Y. Chen, H. Huang, Z. Zhang, V. Lau, andP. Qiu, BSpectrum access for cognitive radionetwork employing rateless code,[ in Proc.Int. Commun. Conf., New York, 2008.IEEE.
[38] N. Nilsson, Principles of ArtificialIntelligence. Palo Alto, CA: Tioga,1980.
[39] H.-M. Huang, BToward a generic modelfor Autonomy Levels for UnmannedSystems (ALFUS),[ in Proc. Perform. MetricsIntell. Syst. (PerMIS’03), Gaithersburg, MD,Sep. 2003.
[40] G. Liu et al., BOn multi-agent collaborativeplanning and its application in RoboCup,[ inProc. 3rd World Congr. Intell. Contr. Autom.,New York, 2000.
[41] P. Stone. (2005). Reinforcement learning forRoboCup soccer keepaway. Sage J. [Online].Available: www.adb.sagepub.com
[42] R. H. Bordini et al., Multi-Agent ProgrammingLanguages, Platforms and Applications.Berlin, Germany: Springer-Verlag, 2005.
[43] B. Bougard, L. Hollevoet, F. Naessens,T. Schuster, C. Ho (Anthony) Ng, andL. Van der Perre, BA low power signaldetection and pre-synchronization enginefor energy-aware software defined radio,[ inProc. SDR Forum Tech. Symp. 2006, Leuven,Belgium. IMEC. [Online]. Available: www.sdrforum.org.
[44] T. Kanter, BA service architecture, test bedand application for extensible and adaptivemobile communication,[ in Proc. PCC’01,New York, 2001.
[45] Proc. 1st Dagstuhl Workshop CognitiveRadio, Dagstuhl, Germany, 2003.
[46] Proc. IEEE Dyn. Spectrum (DySPAN)Conf. 2006 and 2007.
[47] Google, Inc., BLetter to Marlene H. Dortch,Office of the Secretary, FederalCommunications Commission, Re: MHzBands,[ WC Docket 06-150; WC Docket06-129; PS Docket 06-229; WT Docket96-86, May 21, 2007. [Online]. Available:www.fcc.org
[48] F. Fitzek and M. Katz, BCooperation innature and wireless communications,[ inCooperation in Wireless Networks.Dordrecht, The Netherlands: 2006.
[49] F. Fitzek, P. Kyritsi, and M. Katz, BPowerconsumption and spectrum usage paradigmsin cooperative wireless networks,[ inCooperation in Wireless Networks.Dordrecht, The Netherlands: Springer,2006.
[50] A. B. Olsen and P. Koch, BEnergy aware taskallocation in cooperative wireless networks,[in Cooperation in Wireless Networks.Dordrecht, The Netherlands: Springer,2006.
[51] A. Chakrabarti, A. Sabharwal, andB. Aazhang, BCooperative communications,[in Cooperation in Wireless Networks.Dordrecht, The Netherlands: Springer,2006.
[52] P. Mahonen, M. Petrova, and J. Riihijarvi,BCooperation in ad-hoc networks,[ inCooperation in Wireless Networks.
Dordrecht, The Netherlands: Springer,2006.
[53] J. N. Laneman, BCooperative diversity,[ inCooperation in Wireless Networks.Dordrecht, The Netherlands: Springer,2006.
[54] S. Cui and A. J. Goldsmith, BCooperationtechniques in cross-layer design,[ inCooperation in Wireless Networks.Dordrecht, The Netherlands: Springer,2006.
[55] K. Wrona and P. Mahonen, BStability andsecurity in wireless cooperative networks,[in Cooperation in Wireless Networks.Dordrecht, The Netherlands: Springer,2006.
[56] C. Mathur and K. P. Subbalakshmi,BSecurity issues in cognitive radionetworks,[ in Cognitive Networks.New York: Wiley, 2007.
[57] O.-S. Shin et al., BCooperation, competitionand cognition in wireless networks,[ inCooperation in Wireless Networks.Dordrecht, The Netherlands: Springer,2006.
[58] P. Eggers et al., BCooperative antennasystems,[ in Cooperation in WirelessNetworks. Dordrecht, The Netherlands:Springer, 2006.
[59] T. Masden, BCooperative headercompression,[ in Cooperation in WirelessNetworks. Dordrecht, The Netherlands:Springer, 2006.
[60] J. C. H. Lin and A. Stefanov, BCooperativecoding and its application to OFDMsystems,[ in Cooperation in WirelessNetworks. Dordrecht, The Netherlands:Springer, 2006.
[61] Y. Takatori, BCooperative methods forspatial channel control,[ in Cooperationin Wireless Networks. Dordrecht,The Netherlands: Springer, 2006.
[62] Q. H. Mahmoud, Ed., Cognitive Networks.New York: Wiley, 2007.
[63] K. Liebnitz, N. Wakamiya, and M. Murata,BBiologically inspired networking,[ inCognitive Networks. New York: Wiley,2007.
[64] J. Strassner, BThe role of autonomicnetworking in cognitive networks,[ inCognitive Networks. New York: Wiley,2007.
[65] BAn architectural blueprint for autonomiccomputing,[ in Cognitive Networks.New York: Wiley, 2007.
[66] A. Ginsberg, W. Horne, and J. Poston,BThe semantic side of cognitive radio,[ inCognitive Networks. New York: Wiley,2007.
[67] M. Kokar et al., BRoles of ontologies incognitive radios,[ in Cognitive RadioTechnology. Amsterdam, The Netherlands:Elsevier, 2006.
[68] M. Cummings et al., BIEEE 802.21:The leading edge of a larger challenge.[
[69] J. O. Neel, J. H. Reed, and R. P. Gilles,BConvergence of cognitive radio networks,[in Proc. IEEE WCNC, New York, 2004.
[70] Commun. Daily, Dec. 11, 2007.
[71] T. Russell, Signaling System #7, 5th ed.New York: McGraw-Hill, 2006.
[72] M. Sexton and A. Reid, TransmissionNetworking: Sonet and the SynchronousDigital Hierarchy. Norwood, MA:Artech House, 1992.
640 Proceedings of the IEEE | Vol. 97, No. 4, April 2009
[74] J. Davies, Understanding IPv6.Redmond, WA: Microsoft, 2003.
[75] G. Carenini and R. Sharma, BExploringmore realistic evaluation measures forcollaborativefiltering,[ in Proc AAAI 04,Palo Alto, CA, 2004.
[76] A.-B. Garcıa et al., TM Transport BetweenUMTS Base Stations and Controllers:Supporting Topology and DimensioningDecisions. New York: IEEE Press, 2004.
[77] D. Bourse et al., BFP7 E3 project keychallenges,[ in Proc. SDR Forum Tech.Conf. 2007, Denver, CO, Nov. 2007.
[78] S. Zhong, BPerformance evaluation of thefunctional description language in a SDRenvironment,[ in Proc. SDR Forum Tech.Conf. 2007, Denver, CO, Nov. 2007.
[79] R. C. Meier, BCommunication networkproviding wireless and hard-wired dynamicrouting,[ U.S. Patent 0 112 767, 2003.
[80] A. Bravo, BAdvanced positioning andlocation based services In4g mobile-ipradio access networks,[ in Proc. PIMRC,New York, 2004.
[81] N. Bhatia, BPolicy management incontext-aware networks,[ M.Sc. thesis,KTH, Royal Inst. of Technol., Stockholm,Sweden, 2007.
[82] Cantor Telecom, ‘‘Cantor telecomrevolutionizes trading of radio frequencieswith launch of Cantor Spectrum & TowerExchange,’’ New York, Dec. 20, 2004.
[83] F. Capar and F. Jondral, BSpectrum pricingfor excess bandwidth in radio networks,[ inProc. PIMRC, New York, Sep. 2004.
[84] K. Nolan et al., BDynamic spectrum accessand coexistence experiences involving twoindependently developed cognitive radiotestbeds,[ in IEEE DySPAN, New York,Apr. 2007.
[85] K. Nolan, personal communication,Mar. 2008.
[86] D. Maldonado et al., BCognitive radioapplications to dynamic spectrumallocation,[ in Proc. DySPAN, New York,2005.
[87] SAFECOM Program, U.S. Departmentof Homeland Security, BStatement ofrequirements for public safety wirelesscommunications and interoperability,version 1.0,[ Mar. 10, 2004.
[88] S. Li, Univ. of Electronic Science &Technology of China, BContribution forIEEE 802.22 WRAN systems: Sensingscheme for DVB-T,[ Nov. 2007.
[89] S.-G. Jung et al., BDevelopment of URCtesting & certification system,[ in Proc.Int. Symp. Comput. Intell. Robot. Autom.2007 (CIRA 2007), New York, Jun. 2007,pp. 20–23.
[90] T. Kotoku, OMG Working Group Meeting,Tampa, FL, Jan. 2006.
[91] BCognitive vision,[ AAAI Mag., Spring,2006.
[92] J. Mitola, III, BKeynote address,[CROWNCom 2007, New York.
[93] J. S. Olsson, BCombining evidence forimproved speech retrieval,[ in Proc.Doctoral Consortium Workshop(NAACL-HLT 2007), Madison, WI,Apr. 22, 2007.
[94] J. Campbell, BSpeaker recognition:A tutorial,[ Proc. IEEE, Sep. 1997.
[95] BText analytics: On the trail of businessintelligence,[ Nov. 1, 2007. [Online].Available: www.kmworld.com
[96] B. Abu Shawar and E. Atwell, BDifferentmeasurements metrics to evaluate a chatbotsystem,[ in Bridging the Gap: Academic andIndustrial Research in Dialog TechnologiesWorkshop (AACL-HLT), Rochester, NY,Apr. 2007.
[97] J. Stroessner, BAutonomic networks,[ inCooperation in Wireless Networks.Amsterdam, The Netherlands: Elsevier,2007.
[98] M. Cummings et al., BCommercial wirelessmetalanguage scenario,[ in Proc. SDR ForumTech. Conf., Denver, CO, Nov. 2007.
[99] Y. Gao (IBM User Interface Technologies).(2007). BSpeech-to-speech translation.[[Online]. Available: http://www.domino.watson.ibm.com
[100] F. Och. (2006, Apr. 28). Statistical machinetranslation. [Online]. Available: www.googleresearch.blogspot.com
[101] L. Yongquan, BCurrent natural languagecomputing in China,[ Oxford J. LiteraryLinguistic Comput., 1990.
[102] K. Deschacht and M.-F. Moens,BText analysis for automatic imageannotation,[ in Proc. 45th Annu. MeetingAssoc. Comput. Linguistics, Prague,Czech Republic, Jun. 2007.
[103] R. Gopal, BModel based frameworkfor implementing situation managementinfrastructure,[ in Proc. IEEE MILCOM,New York, 2007.
[104] K. Chrisler, BA user-centered approachto the wireless world,[ in Technologiesfor the Wireless Future, R. Tafazolli, Ed.New York: Wiley, 2005.
[105] M. A. Beach, C. M. Tan, and A. R. Nix,BIndoor dynamic double directionalmeasurements,[ in Proc. Int. Conf.Electromagn. Adv. Applicat., Turin, Italy,2003.
[106] J. Stine, BA location-based method forspecifying RF spectrum rights,[ in Proc.DySPAN, 2007.
[107] S. Gandhi et al., BA general frameworkfor wireless spectrum auctions,[ in DySPAN,New York, 2007.
[108] S. Sengupta et al., BAn economicframework for spectrum allocation andservice pricing with competitive wirelessservice providers,[ in DySPAN, New York,2007.
[109] S. D. Jones et al., BCharacterization ofspectrum activities in the U.S. public safetyband for opportunistic spectrum access,[ inDySPAN, New York, 2007.
ABO UT T HE AUTHO R
Joseph Mitola, III (Senior Member, IEEE) received the B.S. degree in electrical
engineering from Northeastern University, Boston, MA, in 1971, the M.S.E. degree from
The Johns Hopkins University, Baltimore, MD, in 1974, and the Licentiate (1999) and
doctorate degrees in teleinformatics from KTH, The Royal Institute of Technology,
Stockholm, Sweden, in 1999 and 2000, respectively.
He is a Distinguished Professor in the School of Engineering and Science and the
School of Systems and Enterprises, Stevens Institute of Technology, Hoboken, NJ, where
his research interests include trustable cognitive systems. Previously, he was the Chief
Scientist of the U.S. Department of Defense (DoD) Federally Funded Research and
Development Center, The MITRE Corporation; Special Assistant to the Director of the
Defense Advanced Research Projects Agency (DARPA); DARPA Program Manager;
Special Advisor to the Executive Office of the President of the United States; and
Technical Director of Modeling and Simulation for DoD. He has also held positions of
technical leadership with E-Systems, Harris Corporation, Advanced Decision Systems,
and ITT Corporation. He began his career as an engineering student assistant with DoD
in 1967.
Mitola: Cognitive Radio Architecture Evolution
Vol. 97, No. 4, April 2009 | Proceedings of the IEEE 641